lazee.dev Jobs API
The dataset behind the lazee.dev job board, over REST and MCP. Open roles pulled straight from 87 company career sites on Greenhouse, Lever, Ashby, SmartRecruiters, Workday, BambooHR, ClanX, Y Combinator, Superteam and Google Careers, tagged by role, level, country and workplace so you can filter them in one request. Refreshed every 12 hours.
Get a key on RapidAPI. Start free.
The Jobs API is sold through RapidAPI. You subscribe there, get your key and are billed there; lazee.dev does not issue keys. The free Basic plan is enough to try every endpoint, and paid plans scale by monthly requests.
Need a field or filter we don't have yet? Tell us and it goes on the list.
Overview
One read-only REST API, JSON in and out, versioned under /v1. It returns the same rows the job board shows: every open role we found on the career sites we track, with the original posting URL. Nothing is user-submitted and nothing is scraped from aggregators; each job comes from the company's own board on its ATS.
- Filter, don't parse. Role, level, region and workplace are already worked out from each posting, so a single query like “senior backend, remote, India” needs no text munging on your side.
- Fresh enough for a job search. The whole dataset is re-pulled every 12 hours. Closed jobs disappear on the next pull instead of lingering for weeks.
- Links you can act on.
urlis the job page on the ATS itself. Every platform listed here is one the lazee.dev extension autofills.
Not included: full job descriptions, salary normalisation, and jobs from boards we do not track yet. See the FAQ.
Data freshness
A scheduled sync runs every 12 hours. It fetches every tracked board in full and brings the dataset in line with what the boards list at that moment:
- New postings are added with
first_seen_atset to the sync time. - Postings whose title, location, pay or URL changed are rewritten and get a new
synced_at. - Postings the board no longer lists are closed: they leave
/jobs, return 404, and are listed in/jobs/closedfor 14 days. - Unchanged postings are left alone, so
synced_attells you when a job last actually changed. - A board that fails to load closes nothing. Its jobs stay as they were, with
source_status: "failed"and the olderlast_checked_atof the last good read.
Closed job or failed refresh?
Every job carries last_checked_at, the last sync that read its board and still found it, and source_status, how that board's last sync went. A job is only ever closed when its board loaded and no longer listed it, so a job missing from your next pull is closed exactly when it shows up in /jobs/closed. A job with source_status: "failed" is unconfirmed, not closed.
/jobs/closedJobs closed in the last 14 days, oldest first. id is the id the job had in /jobs. reason is closed (the board stopped listing it), board_gone (the company took the whole board down) or untracked (we stopped syncing that board). Pass since (ISO 8601) with the last closed_at you stored, plus page and limit. A job that reopens comes back in /jobs under a new id.
{
"data": [
{
"id": "cmh1q7r0b0003pz2m8c5d6e7f",
"platform": "ASHBY",
"company": "Linear",
"title": "Product Designer",
"location": "Remote (US)",
"url": "https://jobs.ashbyhq.com/linear/2b5f1c3e",
"posted_at": "2026-10-02T15:00:00.000Z",
"first_seen_at": "2026-10-03T02:17:00.000Z",
"closed_at": "2026-11-10T14:17:00.000Z",
"reason": "closed"
}
],
"meta": { "page": 1, "limit": 10, "total": 1, "total_pages": 1, "retention_days": 14 }
}/sourcesEvery company board we read, with its status (ok, failed or gone), job count, last good read and, while it fails, failing_since.
{
"data": [
{
"platform": "GREENHOUSE",
"company": "Stripe",
"status": "ok",
"job_count": 512,
"last_checked_at": "2026-11-11T02:17:00.000Z",
"last_attempt_at": "2026-11-11T02:17:00.000Z",
"failing_since": null
}
],
"meta": { "total": 1, "ok": 1, "failed": 0, "gone": 0 }
}What that means for your integration
- A job can be up to 12 hours out of date, and a role closed since the last sync may still appear until the next one.
- Polling more often than every 12 hours buys nothing. Cache responses and re-pull when
meta.last_synced_atchanges. - Use
posted_afterwith your last run'slast_synced_atto pick up only what is new. - Drop closed jobs from your copy with
/jobs/closed?since=your last check; pull at least once every 14 days so none age out first. posted_atis the board's own date when it publishes one. Boards that only show relative dates keep the first value we recorded.- There are no webhooks yet.
Authentication
Every request carries your RapidAPI key in X-RapidAPI-Keyand the API's host in X-RapidAPI-Host. RapidAPI issues the key when you subscribe and shows both values, ready to copy, on the API's Endpoints tab.
GET https://lazee-dev-jobs-api.p.rapidapi.com/v1/jobs?role=frontend
X-RapidAPI-Key: YOUR_API_KEY
X-RapidAPI-Host: lazee-dev-jobs-api.p.rapidapi.com
Accept: application/jsonKeys are for server-side use. Do not ship one inside a browser or mobile app; proxy the call through your own backend instead.
Endpoints
Base URL: https://lazee-dev-jobs-api.p.rapidapi.com/v1. All endpoints are GET and return application/json. The full contract is also available as an OpenAPI 3 spec.
/jobsSearch and filter jobs. Paginated, newest first.
/jobs/{id}One job by id. 404 once its board takes it down, with closed details for 14 days.
/jobs/closedJobs closed in the last 14 days, with when and why.
/sourcesEvery company board with how its last sync went.
/filtersEvery accepted value for role, level, region, workplace and platform, with labels, plus the list of companies.
/statsDataset size, number of companies and platforms, and the last sync time.
{
"data": {
"total_jobs": 1284,
"companies": 87,
"platforms": 10,
"last_synced_at": "2026-10-22T02:00:00.000Z",
"refresh_interval_hours": 12
}
}List and filter jobs
/jobsEvery parameter is optional and they combine with AND. A parameter that accepts several values matches any of them (OR within the parameter). Unknown parameters are ignored; known parameters with a bad value return 400.
| Parameter | Type | Description | Default |
|---|---|---|---|
q | string | Free-text search. Case-insensitive match against the job title, company name and location. | — |
role | enum | Role category worked out from the title. One of the role keys below, or engineering for every engineering role at once. Comma-separate several keys to match any of them. | — |
level | enum | Experience level. One or more level keys, comma-separated. | — |
region | enum | Country or region the job is in. A job with several offices matches every region it mentions. | — |
workplace | enum | remote, hybrid or onsite. Jobs whose board doesn't say have null and never match this filter. | — |
platform | enum | The ATS the job is hosted on. One or more platform keys, comma-separated. | — |
company | string | Exact company name, case-insensitive (for example Stripe). Comma-separate several. | — |
has_compensation | boolean | true returns only jobs whose posting lists a pay range. | false |
salary_min | number | Yearly pay of at least this, in salary_currency: jobs whose pay range reaches it. Other currencies are converted at the day's ECB rate, hourly and monthly pay count as yearly, and jobs that don't list pay never match. 3000000 with salary_currency=INR is 30 LPA. | — |
salary_max | number | Yearly pay of at most this, in salary_currency: jobs whose range starts at or below it. With salary_min, a range. | — |
salary_currency | string | ISO 4217 code of salary_min and salary_max, for example USD, INR or EUR. | USD |
posted_after | date | ISO 8601 date or datetime. Only jobs with posted_at on or after this moment. | — |
sort | enum | posted_desc (newest first) or posted_asc. Ties are broken by id so pages never overlap. | posted_desc |
page | integer | 1-based page number. | 1 |
limit | integer | Results per page, 1 to 100. | 10 |
Response: an array under data plus a meta object.
{
"data": [ /* job objects, see "The job object" */ ],
"meta": {
"page": 1,
"limit": 10,
"total": 1284,
"total_pages": 129,
"last_synced_at": "2026-10-22T02:00:00.000Z"
}
}The job object
/jobs/{id}The same object appears in list results and from the single-job endpoint (wrapped in data). Field names are snake_case; timestamps are ISO 8601 in UTC.
{
"id": "cmh3k9x2a0001qzk8f1g7h4jd",
"platform": "GREENHOUSE",
"company": "Stripe",
"title": "Backend Engineer, Payments",
"location": "San Francisco, CA / Remote (US)",
"workplace": "remote",
"department": "Engineering",
"employment_type": "Full-time",
"compensation": "$180,000 – $250,000",
"salary": { "min": 180000, "max": 250000, "currency": "USD", "period": "year" },
"url": "https://boards.greenhouse.io/stripe/jobs/1234567",
"posted_at": "2026-10-21T09:15:00.000Z",
"first_seen_at": "2026-10-22T02:00:00.000Z",
"synced_at": "2026-10-22T02:00:00.000Z",
"last_checked_at": "2026-10-23T02:00:00.000Z",
"source_status": "ok",
"role": "backend",
"level": "mid",
"regions": ["us"]
}| Field | Type | Description |
|---|---|---|
id | string | Stable identifier for the job. Survives updates to the posting. |
platform | enum | ATS the job is hosted on. One of the platform keys below. |
company | string | Company name as the board publishes it. |
title | string | Job title as posted. |
location | string | null | Location text from the board, untouched. Can list several offices. |
workplace | enum | null | remote, hybrid, onsite, or nullwhen the board doesn't say. |
department | string | null | Board's own department or team label, when it has one. |
employment_type | string | null | Full-time, Contract, Internship and so on, as the board words it. |
compensation | string | null | Pay range exactly as the posting shows it. |
salary | object | null | compensation read as numbers: min and maxper year in the job's own currency(either can be null, as in "up to $180K"), and the period the posting wrote it in. Null when it names no amount. |
url | string | The job page on the ATS itself. Apply there; the lazee.dev extension autofills it. |
posted_at | datetime | The board's posted date. When a board has none, the moment we first saw the job. |
first_seen_at | datetime | When the job first appeared in a sync. |
synced_at | datetime | The last sync that added or changed this job. Unchanged jobs keep their earlier value. |
last_checked_at | datetime | null | The last sync that read the job's board and found the job still listed. Moves every sync, changed or not. |
source_status | enum | null | How the last sync of the job's board went: ok, or failedwhen the board didn't load and the job is kept as it was at last_checked_at. |
role | enum | null | Role key from the title and department. null for roles outside the list (finance, legal, recruiting). |
level | enum | Level key from the title. Titles with no level cue are classed mid. |
regions | string[] | Every region key the location mentions. Empty when none is recognised. |
Filter values
/filtersThese are the keys the filters accept today, generated from the same taxonomy the job board uses. Fetch /filters at runtime rather than hard-coding them; new keys may be added, existing ones are not renamed within /v1.
paramrole
Plus engineering, a shortcut for every key in the Engineering group.
softwareSoftware engineeringfrontendFrontendbackendBackendfullstackFull stackmobileMobileai-mlAI / MLdataDatadevopsDevOps & infrastructuresecuritySecurityqaQA & testinghardwareHardware & embedded
productProduct & programdesignDesignsalesSalesmarketingMarketing & growthsupportCustomer success & solutions
paramlevel
internInternshipentryEntry levelmidMid levelseniorSeniorstaffStaff & principalmanagerManager & above
paramregion
usUnited StatesindiaIndiacanadaCanadaukUnited KingdomeuropeEuropesingaporeSingaporejapanJapanasiaAsia (other)australiaAustralia & NZlatamLatin Americamiddle-eastMiddle East
paramworkplace
remoteRemotehybridHybridonsiteOn-site
paramplatform

GREENHOUSEGreenhouse
LEVERLever
ASHBYAshby
SMARTRECRUITERSSmartRecruiters
WORKDAYWorkday
BAMBOOHRBambooHR
CLANXClanX
YCOMBINATORY Combinator
SUPERTEAMSuperteam
GOOGLEGoogle Careers
{
"data": {
"role": [{ "key": "backend", "label": "Backend" }, ...],
"level": [{ "key": "senior", "label": "Senior" }, ...],
"region": [{ "key": "india", "label": "India" }, ...],
"workplace": [{ "key": "remote", "label": "Remote" }, ...],
"platform": [{ "key": "GREENHOUSE", "label": "Greenhouse" }, ...],
"company": ["Airbnb", "Anthropic", "Stripe", ...]
}
}Pagination
Results are paged with page and limit. The default page size is 10 and the maximum is 100. meta.total is the number of jobs matching your filters and meta.total_pages is the last page you can request; asking for a page past it returns an empty data array, not an error.
Ordering is by posted_at then id, so paging through a result set between two syncs is stable. A sync in the middle of a long crawl can shift rows; compare meta.last_synced_at across pages and restart if it changed.
Errors
Errors use standard HTTP status codes and a JSON body with a machine-readable code and a human message. Validation errors also name the offending param.
HTTP/1.1 400 Bad Request
Content-Type: application/json
{
"error": {
"code": "invalid_parameter",
"message": "level must be one of: intern, entry, mid, senior, staff, manager",
"param": "level"
}
}| Status | code | Meaning |
|---|---|---|
| 400 | invalid_parameter | A query parameter has a value outside its allowed set or format. The response names the parameter. |
| 401 / 403 | — | The RapidAPI key is missing or invalid, or you are not subscribed. Sent by RapidAPI with its own message body. |
| 404 | not_found | No job with that id. Jobs taken down by their board are deleted on the next sync and 404 afterwards. |
| 429 | — | Over your plan's rate limit or quota. Sent by RapidAPI; back off and retry. |
| 500 | internal_error | Something failed on our side. Safe to retry with backoff. |
Rate limits and pricing
Plans, quotas and prices live on the RapidAPI listing: a free Basic plan for prototyping and paid plans by monthly request volume. RapidAPI enforces rate limits per key and answers 429 when you go over.
Because the data only changes every 12 hours, most integrations need far fewer requests than they first assume. One full pull per sync, filtered to what you need, is usually enough.
Use of the API is subject to the Jobs API terms.
Examples
Senior backend or full-stack roles, remote, in India:
curl "https://lazee-dev-jobs-api.p.rapidapi.com/v1/jobs?role=backend,fullstack&level=senior®ion=india&workplace=remote&limit=10" \
-H "X-RapidAPI-Key: YOUR_API_KEY" \
-H "X-RapidAPI-Host: lazee-dev-jobs-api.p.rapidapi.com"Engineering roles mentioning “payments”, posted this month, with a listed pay range:
const params = new URLSearchParams({
q: "payments",
role: "engineering",
has_compensation: "true",
posted_after: "2026-10-01",
limit: "50",
});
const res = await fetch(`https://lazee-dev-jobs-api.p.rapidapi.com/v1/jobs?${params}`, {
headers: {
"X-RapidAPI-Key": process.env.RAPIDAPI_KEY,
"X-RapidAPI-Host": "lazee-dev-jobs-api.p.rapidapi.com",
},
});
if (!res.ok) throw new Error(`Jobs API ${res.status}`);
const { data: jobs, meta } = await res.json();
console.log(`${meta.total} jobs, refreshed ${meta.last_synced_at}`);
for (const job of jobs) console.log(job.company, "—", job.title, job.url);Entry-level and intern roles in the US on Ashby or Lever:
import os, requests
resp = requests.get(
"https://lazee-dev-jobs-api.p.rapidapi.com/v1/jobs",
params={"platform": "ASHBY,LEVER", "level": "entry,intern", "region": "us", "page": 1},
headers={
"X-RapidAPI-Key": os.environ["RAPIDAPI_KEY"],
"X-RapidAPI-Host": "lazee-dev-jobs-api.p.rapidapi.com",
},
timeout=15,
)
resp.raise_for_status()
body = resp.json()
for job in body["data"]:
print(job["company"], "-", job["title"], job["url"])
print("page", body["meta"]["page"], "of", body["meta"]["total_pages"])MCP server
The same data works inside AI apps such as Claude, ChatGPT, Cursor and VS Code, so they can search jobs for you in plain language: “find remote senior backend roles in India posted this week”. There are two ways to connect.
1. lazee.dev MCP server: sign in, no key
Add https://lazee.dev/mcp to your AI app and sign in with your lazee.dev account. 100 calls a month free, 5,000 on Pro. In Claude, add it straight from the connector directory. Setup for each app.
2. RapidAPI's MCP server: your RapidAPI plan, for higher volumes
RapidAPI also serves this API as an MCP server, billed on your RapidAPI plan like any other request, so paid plans take you past the lazee.dev limits. Add this to your client's MCP config (for Claude Desktop, claude_desktop_config.json; for Cursor, .cursor/mcp.json) and replace YOUR_RAPIDAPI_KEY with your key. It needs Node.js 18+ for npx.
{
"mcpServers": {
"RapidAPI Hub - lazee.dev Jobs API": {
"command": "npx",
"args": [
"mcp-remote",
"https://mcp.rapidapi.com",
"--header",
"x-api-host: lazee-dev-jobs-api.p.rapidapi.com",
"--header",
"x-api-key: YOUR_RAPIDAPI_KEY"
]
}
}
}The config holds your key, so keep it out of shared repos.
FAQ
Where do I get access?
On RapidAPI, where you subscribe, get a key and pay. Start on the free Basic plan and upgrade when you need more requests.
Does it include the job description?
Not yet. Each job carries the URL of the original posting, where the description lives. Descriptions are the most requested addition, so tell us if you need them and how you would use them.
Can I filter by salary?
Yes. salary_min and salary_max take a yearly amount in salary_currency, and match jobs in any currency, converted at the day's ECB rate. Each job's salary has its pay as numbers, per year in its own currency. Only jobs that list pay can match (about a fifth of the board), and conversions are approximate. has_compensation=truenarrows to jobs that list a range at all. Sorting by pay isn't offered.
How accurate are role, level and region?
They are derived from the title, department and location with rules tuned on the boards we track, not read from a structured field, so a small share of jobs are tagged imperfectly. A title with no level cue is classed mid; a role we don't categorise (finance, legal, recruiting) has role: null.
Can you add a company or a board?
Yes, if it is hosted on one of the supported platforms. Send the company's careers URL through the feedback form.
Can I use the job board instead?
You can browse and filter the same dataset for free on the job board. For programmatic access use the API; please do not scrape the board.
